Circuit to start with worst case line conditions in terms of inrush

[CopperCone]

So as part of my isn related threads i was interested in designing some sort of circuit which switches in ac line voltage in the worst place of the cycle.

What is the best way to do this? The idea is so it can reliably trigger a inrush to see how inrush emc looks like and to be able to do something like test the input circuitry for 1000 worst condition starts and check for component degredation etc

I thought a path towards doing this would be to know line frequency, then use some kind of mains connected zero crossing detector to connect to a delay circuit which would switch on a triac when peak mains ac is anticipated or at any point in its cycle since its frequency stable.

Is a triac a good choice for this? Or do you need some kind of transmission gate? What is a good idea for a reliable zero crossing circuit to act as a trigger?

Otherwise its turn something on 50 times and see if it does what you want

More difficult if the circuit you are using has a soft start but its a start

[Benta]

"Worst Case" depends on the load you are switching. For a resistive load, worst case would be at maximum voltage. But for a transformer, maximum voltage would be best case and zero voltage switching worst case. And for mixed loads (eg, transformer/rectifier/filter cap), worst case is probably a case for trial-and-error.

A triac is a good choice, but it needs to be large. I'd suggest two SCRs or two IGBTs for maximum ruggedness.

And as said, you need to make your circuit programmable somehow to cover a full half-cycle (or a full cycle if your load is asymmetric).

[Ian.M]

Additionally, for repeatable results with a transformer or inductive load, you need to remember the polarity of the last half-cycle before switch-off, so you can switch-on during the same polarity half cycle, so any residual magnetisation aids saturation rather than opposes it or randomly ads or opposes it.

[Siwastaja]

Indeed, the "worst case" really depends and it's hard to predict exactly. You can end up doing an optimistic assumption while thinking you are testing the worst case! While you can, with enough experience and knowledge, actually analyze the worst case conditions and simulate only them, it defeats the main purpose of testing: to verify that the knowledge and analysis was indeed correct!

The "state" of the system affects the worst case (for example - at what phase was the power removed, affecting residual magnetism), as well as the delay between the tests.

Instead of trying to find the single "worst" condition and sync to it exactly, I'd come up with automated way to generate connections/disconnections and log the results. Then, you can throw it with a truely random test sequency which will find you the actual worst case operating condition in around 100 cycles.

The test would be out of sync from the phase, and include long delays as well as very short delays.

In some designs, such as those using NTCs to provide inrush limiting, short-delay toggling or input power bouncing requires very careful analysis (or, the designers just ignore the issue).

[CopperCone]

Yes this whole process is a rabbit hole i would like to explore.

I am worried that using a semiconductor switch will not offer the best results because when a relay or switch is toggled there is a spark gap effect. I am still studying it but i believe that the spark will act as a kind of pulse sharpener so you will get a specific burst profile of high frequency interference that should not be present with semiconductor switching.

Now this is problematic because the pulse sharpening behavior of switches, contactors, relays or plugging it into the wall while turned on will all be different i believe it depends on electrode geometry, materials choice of arcing terminals and atmospheric conditions specifically humidity, temperature may be very minor close to ambient.

I thought to use a triggered spark gap which would have the plasma channel shorted by a solid state switch within something like 1/8 of a cycle may be an option but this of course imposes my own sparkgap interference standard on the device being tested and the conditions may not be a good replication of the sparking that occurs during mechanical throw.

Does commonly used power factor correction circuitry have provisions for inrush current supression and stuff? Are there accepted standards or is it the wild west?

I am interested in complex systems interactions personally, so this is for my own interest rather then trying to meet some kind of commercial expectations.

[CopperCone]

Btw i like my idea of using a triggered gap which is shunted by the triac or scr or igbt because the spark can pass a good portion of the insane current levels (i want to set up this test with 10ga wire a few feet away from the distribution box on its own breaker so i can get the most out of my pole transformers peak and lower inductance.

Any parts suggestions for me? I know nothing of power scr, nothing about triac and a bit about igbt. I only used signal level scrs for high reliability stuff and i just never ventured into the triac, sidac, diac, sidactor, etc domain

I also need to look at voltage drop of the sparcgap arc to make sure its ok.

I want to test things on the worst possible setting so the rise time needs to be high and it needs to be able to be low impedance so i dont get a reduced idea of what is going on.

The use of this device is not restricted to the isn, since it has significant inductance, so it needs to be robust and the power section needs to be unisolated.

What sort of circuit would be used with igbt ?
I figure you would need some kind of diode steering circuit?

What kind of triac would you use for this? What kind of scr circuit would you use?

Also is some kind of transmission gate possible using nmos and pmos hv? Not sure if they even have pmos hv. Any benefits to this?

With my sparkgap idea i figure i would need to measure current and design it to switch in the transistor bypass when the current drops to acceptable levels? Though this circuit seems hairy.it would need to be before the arc is extinguished though.

[2N3055]

Well, you can always try a chattering relay....

[CopperCone]

Eh thats pretty bootleg and you cant do advanced diagnostics like

1 run bad power up condition
2 Interface to equipment on rs23w to see if memory initalized properly

I did find a elevator relay last week that i will make a relay chattering circuit with though

[CopperCone]

God damn need for rogawski coils is making me uncomfortable with this design.

I want to know how much the switch spark effects the measurement to know if i should get into crazy shit with discharge tubes.

I figure if i use a regular current probe it will saturate so i cant measure the emp.

Do they have some kind of probes that will be immune to fields under a few khz that can take the massive inrush but give meaning ful measurements on the hf noise?

Not sure where i am going with this i am kinda thinking i want to avoid building stuff without its own isolation transformer and some kind of zero crossing trigger circuit of its own simply because putting the extra power electronics into my designs is cheaper then failure given that i would have to make enclosures and cooling system and all that shit

And it seems that for properly designed modern electronics you would need to hack the pushbuttong switch to the 'worst case' circuit with some kind of time delay that you measure on the circuit being tested which should have provisions for this anyway

It seems of limited use but i guess i would like to still make it to evaluate the survivability of electronics which have failed pfc/inrush control but it seems this is only of interest to the fanatic at this point.

With out a sparkgap sharpening effect happening to mangle the noise it seems only like some sort of magnetic field tester

Isuppose the utility can be increased if you can quickly trigger a impulse generator tosum into the switching junction but the noise profile wont be the same

I supose that the iec standards kind of give you an idea of what to look for.

But due to switch variability and how the noise profile changing with switch type, other conditions, it may be better to just leave provision for interface with some kind of adjustable impulse generator in order to test the circuit with many different options ratber then relying on the sparkgap in the switch to act as the sharpener.

Now i just need to evaluate the triac, scr or igbt choice with peak currents, determine the possible benefit of using a nonsharpening triggered spark gap to increase switch survivability

I figure the circuit will be using a bandpass filtered 60hz signal to connect to a zero crossing comparator which will trigger variable one shot time delays to the semiconductor and possibly gas switch.

If the gas switch has merit then i suppose you would want a fucking rogawski coil to measure the current to a safe level for the semiconductor to switch, but it must switch before the arc is extinguished.

So other then power switch evaluation i neeed to determine arc voltage drop.

[CopperCone]

Hmm then there is also the question of injecting heavy impulses into a power company feed

I guess you would need some small value large physicsl size air inductors to maintain some discression while testing so you dont have the gestapo show up..

Or a 250 amp isolation transformer that can pass the surge. Or a smaller torroid.

But then i think you can get some dirty behavior.

At least you dont need to stabilize it for measurement though. Its a sledge hammer test

[Kleinstein]

An inductor or isolation transformer would not help very much in isolation the grid from the inrush current. There can be some reduction in the inrush current, but the current still comes from the grid. In an 1:1 transformer the current is still nearly the same on both sides. Starting with a higher voltage (e.g. 500 V range) might help though as those higher voltages a re usually used for higher power. 

Essentially the only way to isolate the pulse from the grid would be to have a kind of local generator, like the old motor - generator sets to generate higher frequencies, but used to make 60 Hz again. A large motor with a flywheel could also stabilize the net. Still it's complicated as it needs to be rather large (e.g. 10 kW range)  to really have an effect and that size motors also don't start easy on a 10 A fuse.

For the very short pulse part a reasonable sized capacitor (e.g. a few 10 µF)  could help a little, though it would initially have some capacitive current.  Depending on the circuit the inrush currents can be quite high - possibly up to the limits of triggering an 30 A fuse, so maybe in the 100s of A. So make sure the fuses have sufficient brake capability.

[CopperCone]

No no I meant to isolate something like a spark gap impulse generator from the grid so you don't pump giant MHz spikes into it, I want the surge current to max out their pole transformer, for the purposes of the test

My idea is to plug an impulse generator into my mains trigger thing, and trigger the impulse generator to spike it during the inrush to see if it has an effect and to simulate possible switch related noise

I don't want to plug a spark gap into the outlet witout some kind of filter because it might interfere with neighbors, power line communications, whatever else is going on